OBJECTIVETo explore the correlation between incidence of atrial fibrillation (AF) and thyroid dysfunction.

METHODSPatients with stable angina pectoris with thyroid function test results hospitalized at Fuwai Hospital from 2011 Jan to 2011 Dec were included in this analysis (n = 2 541). General clinical data and related biochemical parameters were analyzed. We divided patients into 5 subgroups according to TSH levels: <0.55 mIU/L (n = 105), 0.55-2.49 mIU/L (n = 1599), 2.50-4.77 mIU/L (n = 621), 4.78-9.99 mIU/L (n = 180), >10.00 mIU/L (n = 36).

RESULTSA total of 157 patients were diagnosed with AF (6.8%). (1) Compare to stable angina pectoris patients without AF, stable angina pectoris patients with AF have older age (P < 0.001), higher proportion of female (P = 0.04), uric acid (P < 0.001), NT-proBNP (P = 0.001), larger left atrial diameter (P < 0.001), left ventricular end diastolic diameter (P < 0.001) and lower LVEF (P = 0.038), FT3(P = 0.002), TT3 (P < 0.001). (2) When TSH levels were less than 0.55,0.55-2.49, 2.50-4.77, 4.78-9.99 mIU/L and greater than 10.00 mIU/L, the incidence of AF were 7.6% (8/105) , 5.7% (91/1 599), 7.9% (49/621), 9.4% (17/180) and 22.2% (8/36), respectively. Both a high and a low TSH level were associated with an increased incidence of AF. After adjustment for common risk factor (age, gender and so on) , stepwise multiple logistic regression analysis revealed that TSH levels were significantly related with the incidence of AF. Compared to patients with TSH 0.55-2.49 mIU/L, the adjusted odds ratio of AF for TSH < 0.55, 2.50-4.77, 4.78-9.99, >10.00 mIU/L were 1.37 (95%CI 0.65-2.90, P = 0.415), 1.42 (95CI 0.99-2.04, P = 0.057), 1.73 (95%CI 1.01-2.97, P = 0.048), 4.74 (95%CI 2.10-10.69, P < 0.001), respectively.

CONCLUSIONOur results show that incidence of AF increases in proportion to TSH level in patients with stable angina pectoris.